EP0237404A1 - Method of processing sum and difference signals in a monopulse radar for estimating the parasitic phase between these signals caused by the HF circuits forming sum and difference channels - Google Patents

Abstract

The method according to the invention, applied for example to the estimation of the parasitic phase Ψa between the sum Σa and difference Δa in azimuth channels, consists in calculating the expressions S = Σa. Σe * and D = Δa. Σe * for different successive measurements made by the radar, then calculating the expression [S (k + 1) - S (k-1)]. D * (k) for different successive instants k-1, k, k + 1, and to carry out an averaging of the result of this expression for different instants k, it being understood that the radar antenna scans in azimuth and that the radar wave emitted is circularly polarized, the azimuth and elevation monopulse receivers respectively receiving only one of the types of circular polarization: right or left. Application to tracking cameras of the monopulse type.

Description

The present invention relates generally to the radar field and relates more particularly to a method of processing signals from a radar of the monopulse type, with a view to estimating the parasitic phase introduced between these signals by the microwave training circuits. sum and difference paths.

It is known that a monopulse type radar allows the angular location of targets by processing different signals received simultaneously from these targets and corresponding to different beam directivities.

Thus, as shown very schematically in FIG. 1, by providing two sources respectively in the azimuth plane (1a, 2a) and in the elevation plane (1e, 2e), there are two signals available for each of these planes are we can form, by microwave means (3,4), the sum Σ and the difference Δ, which amounts to having an antenna with a beam corresponding to channel la and another beam corresponding to channel A, and thus allows angular localization of the targets in a simple manner.

avec:

• s(θ _i) : gain de la voie somme à l'angle θ_i ;
• d(θ _i) : gain de la voie différence à l'angle θ_i ;
• a_i: amplitude relative au iemeréflecteur ;
• φ_i déphasage introduit par le jèmeréflecteur ;
• d_i: distance entre le radar et le ièmeréflecteur ;
• Ψ: phase parasite introduite entre les voies Σ et A par les circuits hyperfréquence de formation des voies somme et différence.

It is also known that, the re-radiation of a target being the sum of the waves reflected by each of the reflective elements which compose it, the signals S and D received respectively on the sum and difference channels, in azimuth or in elevation, have for representation complex:

with:

• s (θ _i ): gain of the sum channel at angle θ _i ;
• d (θ _i ): gain of the channel difference at angle θ _i ;
• a _i : amplitude relative to the iemoreflector;
• φ _i phase shift introduced by the jem reflector;
• d _i : distance between the radar and the ith reflector;
• Ψ: parasitic phase introduced between channels Σ and A by the microwave circuits for forming the sum and difference channels.

où D^* désigne la quantité complexe conjuguée de D, et où :

It is also known that, in the particular case of the elevation plane, the parasitic phase Ψ between the sum and difference channels can be evaluated in the following manner. Let P = S x D ^*

where D ^* denotes the complex conjugate quantity of D, and where:

pouvant être considérée comme aléatoire, on a :

où E désigne l'opérateur espérance mathématique.If the average value of P is calculated over different distance from the radar, 2d .. the phase φ _ij - w

can be considered random, we have:

where E denotes the operator mathematical expectation.

The term "a" is real, but is a priori of unknown sign because of the factors d (θ _i ) which can be positive or negative. Indeed if we refer to Figure 2 which represents the shape of the sum and difference diagrams, Σ and Δ, in the azimuth plane or in the elevation plane, the sum diagram presents a maximum in the direction of the axis of the antenna, while the difference diagram A has on the contrary a minimum in this direction, and is, on either side of this direction, either in phase (Δ +) or in phase opposition (A-) with Σ.

The averaging of P then allows a priori to estimate the phase Ψ only to the nearest sign.

Now it turns out that in the elevation plane, this sign ambiguity can be removed relatively simply because the sign of d (θ _i ) is then linked to the rank of the distance box considered, being negative for the closest distance boxes and positive for the most distant distance boxes. By averaging P over the different distance boxes, we then obtain, after correction of the sign according to the distance box considered, an unambiguous estimate of the phase Ψ.

The present invention as defined in the claims relates to a method for estimating the parasitic phase Ψ making it possible to remove these ambiguities of sign also in the azimuth plane, and the principle of which is also applicable to the estimation of the parasitic phase Ψ in the elevation plane, as well as in the estimation of the differential phase δΨ between the azimuth and elevation planes, this latter possibility also constituting, as will be seen below, the basis of another estimation method of the parasitic phase in the azimuth plane.

• - la figure 1 montre schématiquement la formation des voies somme et différence en azimut et en élévation dans un radar de type monopulse ;
• - la figure 2 montre la forme des diagrammes somme et différence d'un radar monopulse dans le plan azimut ou dans le plan élévation ;
• - la figure 3 illustre le balayage de l'antenne radar, suivant l'invention ;
• - la figure 4 est un schéma synoptique du traitement effectué suivant l'invention ;
• - la figure 5 est un schéma du détecteur de produit utilisé dans le traitement suivant l'invention.

Other objects and characteristics of the present invention will appear more clearly on reading the following description of exemplary embodiments, given in relation to the attached drawings in which:

• - Figure 1 schematically shows the formation of sum and difference channels in azimuth and elevation in a radar of the monopulse type;
• - Figure 2 shows the form of the sum and difference diagrams of a monopulse radar in the azimuth plane or in the elevation plane;
• - Figure 3 illustrates the scanning of the radar antenna according to the invention;
• - Figure 4 is a block diagram of the processing carried out according to the invention;
• - Figure 5 is a diagram of the product detector used in the treatment according to the invention.

The invention is used on a radar system with coherent or non-coherent emission, circularly polarized, capable of receiving and discriminating right circular polarization and left circular polarization. To fix the ideas in what follows, the azimuth channels will receive the right polarization while the elevation channels will receive the left polarization; the reverse would also be possible.

où les indices D et G désignent les voies de réception des polarisations droite et gauche, et les indices a et e les dimensions angulaires, azimut et élévation, et où ϕ_t désigne la composante de phase non cohérente qui change d'impulsion à impulsion.Thus by referring to a hypothetical absolute phase the sum signals Σa and Σ _e , respectively in azimuth and in elevation, and the difference signals A _a and Δ _e , respectively in azimuth and in elevation, are written:

where the indices D and G denote the reception channels of the right and left polarizations, and the indices a and e the angular dimensions, azimuth and elevation, and where ϕ _t denotes the non-coherent phase component which changes from pulse to pulse.

From the sum and difference signals Σa, Σe, Δ _{a '} Δe, the following expressions are calculated:

A unit of time being fixed small enough so that the phase shift between channels can be considered as constant, we have at an instant k:

It will be noted that the computation of the expressions S and D allows, inter alia, to be freed from the non-coherent phase component ϕt.

avec

We then calculate the expression: [S (k + 1) - S (k-1)] D * (k) where k-1, k and k + 1 denote three successive instants. We have :

with

In the expression Σ b _n eiϕ n, the factors b _n are assigned a phase term ϕ _n which, which can be considered as random, will disappear by averaging the expression: [S (k + 1) S (k -1)] D ^* (k) on several successive measurements of Σa, Δa, Σe, Δe taken by the radar and corresponding to several successive instants k.

The notion of radar measurements relates to the notion of distance and pulse boxes. For a given distance cell, the averaging is then carried out on several successive pulses, and the averaging is then carried out on several distance cells. Thanks to the double polarization the factors not affected by such a phase term are limited to the expression "a" for which there is the equality i = I and j = m. All the other cases (such as i = j or m = I) are indeed excluded because they would correspond to individual reflectors which would return the two types of polarization: right and left, that is to say which would be at the same time of the dihedral type (even number of successive reflections, and of the lukewarm type (odd number of successive reflections).

The invention therefore makes it possible to conserve all the angular information included in the signals received, while eliminating the non-coherence of the transmission, by referencing them to the response to the same pulse of different reflectors. This is achieved by the combination of the calculation of expressions S and D and the double polarization.

/ During the previous development it was made the important assumption that between the instants k-1 and k + 1 the elevation angles θ _e (in this case θ _G since we consider in this example the case where the azimuth channels receive the right polarization while the elevation channels receive the left polarization) are invariant, which allows them to be put into common factor in the expression "a". This assumption is reasonable in the case of a fixed radar or an on-board radar if the time unit is small enough so that the advancement of the platform does not change the geometry of the system.

The invention is also used in an azimuth scanning radar system, the antenna rotation speed (in azimuth) being fixed at a value such that the lobes Δ - and Δ + at the instant k coincide respectively with the lobe Σ at time k - 1, as shown in FIG. 3.

Under these conditions the product [Sa (k + 1) (θ Di) -s _{a (k-1)} (θ Di)] da (k) (θ Di) is positive for any elementary reflector "i". Indeed, if we consider for example an elementary reflector which is at the instant k on the lobe A +, we then have d _a ( _k ) which is positive, s _{a (k + 1)} which is also positive; it would be the same for an elementary reflector which would be at the instant k on the lobe Δ-.

The rotational speed of the antenna in azimuth is not critical and could be chosen differently. The choice proposed however seems optimal in that it gives the largest values of the difference (Sa ( _k + 1) - S _{a (k-1)} ).

The factor "a" is therefore written as the sum of a large number of positive reals. On the contrary, the factors bn are assigned a phase term ϕ _n which can be considered as random.

1 -

₃ dont la variance variera inversement au nombre de mesures effectuées, ce nombre de mesures étant fixé a priori ou adapté à la précision désirée par software.The examination of several successive measurements of [S (k + 1) - (S (k-1)] D ^* (k) allows by averaging to estimate the phase correction

1 -

₃ , the variance of which will vary inversely with the number of measurements carried out, this number of measurements being fixed a priori or adapted to the precision desired by software.

The adaptation of the period of repetition of the measurements carried out by the radar (corresponding to the notion of distance and pulse boxes) and of the period of repetition of the instants k, that is to say of the scanning speed of the antenna can be made in various ways, for example by suitably choosing the repetition period of the pulses, or else by performing post-integration on several pulses.

FIG. 5 shows a diagram of the product detector used in the treatment carried out according to the invention to estimate the parasitic phase between the sum and difference channels in azimuth.

au moyen de deux mélangeurs 5 et 6 dont l'un reçoit d'une part le signal Σ_e et d'autre part, soit le signal Σa, soit le signal Δa. Les signaux Σa, Δa et Σ_esont par ailleurs transformés en signaux à moyenne fréquence au moyen d'étages de changement de fréquence 8 et 9.The signals Σa, A _a and Σ _e are applied to the input of this product detector which calculates the expressions:

by means of two mixers 5 and 6, one of which receives the signal Σ _e on the one hand and the signal Σa or the signal Δa on the other hand. The signals Σa, Δa and Σ _e are moreover transformed into medium frequency signals by means of frequency change stages 8 and 9.

It should be noted that S and D are not used at the same times, since only the product [S (k + 1) - S (k-1)] D ^* (k) matters. It is therefore possible to use only one product detector, addressed successively, by means of a switch 10 by the channels Ea and Ec then Δ _a and Σe.

The quantities S and D are complex quantities which can be written in the form: I + j Q where I and Q respectively designate their real part and their imaginary part; the quantities I and Q are obtained on two outputs of the product detector.

The other steps of the process can be carried out using suitable software.

In the foregoing, the invention has been described for the estimation of the parasitic phase between the sum and difference channels in azimuth. The principle of the invention is nevertheless applicable to the estimation of the parasitic phase between the sum and difference in elevation channels.

In this case we would calculate the expressions:

The constraint on the scan would then become a constraint on the elevation scan, and the assumption that between the instants k-1, k and k + 1 the azimuth angles are invariant should also be made.

As we will see now, the principle of the invention is also applicable to the estimation of the differential parasitic phase Ψ _a - Ψ _e , where Ψ _a and Ψ _e denote respectively the parasitic phase in azimuth and in elevation. The estimation of this differential parasitic phase may be of interest for the following reasons.

In the case in particular of a radar carried by a missile whose supersonic speed decreases during flight to subsonic values, the significant temperature differences between the start and the end of the mission cause a variation of the parasitic phases in azimuth and in elevation.

However, the antenna scanning constraints described above mean that the estimation of a phase correction factor cannot be carried out during the entire period of use and processing of the radar.

Starting from the observation that the four receivers (two in azimuth, and two in elevation) are physically close to each other and thermally interdependent, therefore that the variations of Ψ _a and Ψ _e will be similar, it suffices then to evaluate Ψ _a and Ψ _e during a so-called calibration phase (prior to any operation in search or pursuit mode) and to calculate the difference Ψ _a - Ψ _e which will remain constant throughout the duration of the flight. In the tracking phase it is always possible, as recalled in the introduction, to calculate Ψ _e by an evaluation mode which offers no constraint on antenna position or frequency. It is then possible to deduce Ψa from it.

However, a successive evaluation of Ψ _a and Ψ _e during the calibration phase risks inducing an error on Ψ _a - Ψ _e since the temperature may have changed between the two evaluations. In addition, the calibration phase has a significant duration compared to the total duration of the mission of a short-range missile.

The parasitic phase estimation method according to the invention makes it possible to directly calculate the differential phase Ψ _a - Ψ _e without going through the successive calculations of Ψ _a and Ψ _e , which makes it possible to overcome the risk of error previously. cited.

où les indices D et G désignent les voies de réception des polarisations droite et gauche (en supposant toujours que les voies azimut reçoivent la polarisation droite, et les voies élévation la polarisaton gauche) et les indices a et e les dimensions angulaires, azimut et élévation.

• a_i est l'amplitude du iéme réflecteur individuel ;
• φ est le déphasage introduit par le ième réflecteur ;
• d_i est la distance radar-ième réflecteur ;
• ϕt est la composante de phase non cohérente qui change d'impulsion à impulsion ;
• Ψ₁, Ψ₂, Ψ₃, Ψ₄ sont les phases parasites introduites sur les quatre voies de réception.

We now describe this process, applied to the estimation of the differential phase Ψ _a Ψ _e . As before, the signals received by the four deviation channels are written:

where the indices D and G denote the reception channels of the right and left polarizations (always assuming that the azimuth channels receive the right polarization, and the elevation channels the left polarization) and the indices a and e the angular dimensions, azimuth and elevation .

• a _i is the amplitude of the ith individual reflector;
• φ is the phase shift introduced by the ith reflector;
• d _i is the radar-th reflector distance;
• ϕt is the non-coherent phase component which changes from pulse to pulse;
• Ψ ₁ , Ψ ₂ , Ψ ₃ , Ψ ₄ are the parasitic phases introduced on the four reception channels.

Alors : [S(k+1) -S(k-1)] D^* (k) = e-j(Ψ2-Ψ1 + Ψ3-Ψ4)[a +

bn e jφn] où les phases φn peuvent être considérées comme des phases aléatoires et où : a =

[s_a(k + ₁₎(θ _Di) -s_a(k-1)(θ _Di)] d_a(k)(θ _Di)s_e(θ _Gj)d_e(θ _Gj)a² _Dia²GjUsing the same method as before, a product detector calculates:

So : [S (k + 1) -S (k-1)] D ^* (k) = ej (Ψ2-Ψ1 + Ψ3-Ψ4) [a +

bn e jφn] where the phases φn can be considered as random phases and where: a =

[s _a (k + ₁₎ (θ _Di ) -s _{a (k-1)} (θ _D i)] d _{a (k)} (θ _Di ) s _e (θ _Gj ) d _e (θ _G j) a ² _Sun a ² Gj

Now the speed of rotation in azimuth of the radar was chosen in such a way that the product: [s _{a (k} + ₁₎ (θ _Di ) s _{a (k-1)} (θ _Di ] d _{a (k)} (θ _D i) is positive for all the elementary reflectors. Furthermore s _e (θ _G j) is positive, a ² Dia ² Gj is positive, and de (θ _G j) is negative for the closest distance boxes, and is positive for the most distant distance boxes.

The ambiguity of the sign can therefore be removed and corrected in such a way that the factor "a" is the sum of a large number of positive reals.

1 -

2 -

3+

4 c'est-à-dire Ψa - Ψe.The examination of several successive measurements of [S (k + 1) -S (k-1)] D ^* (k) allows by averaging to estimate the phase correction

1 -

2 -

3+

4 i.e. Ψa - Ψe.

The average is calculated over several distance boxes and several pulses so as to obtain the desired precision.

Claims (3)

1. Method for processing the sum Σ and difference A signals, in azimuth: Ea, Δa, and in elevation: Σ _e , A _e , of a monopulse type radar, with a view to estimating the parasitic phase Ψ introduced between these signals by the microwave circuits (3, 4) for forming the sum and difference channels, characterized in that it consists of: - calculate for each measurement carried out by the radar the expressions S and D, with

or

or

(where the symbol ^* denotes the complex conjugate quantity), depending on whether one wishes to estimate the parasitic phase Ψa in azimuth in the first case, or the parasitic phase Ψe in elevation in the second case or the differential parasitic phase δΨ = Ψ _a - Ψ _e in the third case, it being understood that the transmitted radar wave is circularly polarized and that the azimuth (1a, 2a) and elevation (1e, 2e) monopulse receivers respectively receive only one of the types of circular polarization: right or left, - calculate the expression [S (k + 1) -S (k-1)] D * (k) it being understood that the radar antenna scans during the tenps in azimuth in the first and in the third case, in elevation in the second and that k-1, k, k + 1 denote three successive instants such that the difference s _{(k + 1} (θ) - s _(k-1) (θ) has the same sign as d _(k) ( θ), where s _{(k + 1)} (θ) and S _(k-1) (θ) denote the gain of the sum channel at angle (θ) respectively, in azimuth in the first and in the third case, in elevation in the second case, respectively at times k + 1 and k-1, and where d _(k) (θ) denotes the gain of the difference path at angle θ and at time k, in azimuth in the first and in the third case, in elevation in the second case; - calculate the average value of the result of the previous expression for several successive instants corresponding to several measurements carried out successively by the radar, which makes it possible to obtain an expression of the form ejΨ. a, where Ψ is the parasitic phase sought and "a" a real number of determined absolute value and sign. 2. Method according to claim 1, characterized in that the differential parasitic phase δΨ is estimated in a calibration phase prior to the use phase of the radar, and implementing an antenna scan, the parasitic phase Ψ _has in azimuth then being deduced, during the radar use phase, from the differential parasitic phase δΨ thus estimated during the calibration, and of the parasitic phase in elevation Ψ _e estimated during the phase of use of the radar, without it being necessary to implement an antenna scan, by averaging the expression Σ _a . Δa * on several successive measurements made by the radar. 3. Method according to one of claims 1 and 2, characterized in that only the product [S (k + 1) - S (k-1)] D ^* (k) important, the expressions S and D are calculated at means of a single device used successively for the calculation of S and D.

Images